Tweet
Juan, thank you for responding and burning this down for us. And Enzo, thank you for taking the lead on this. And let me say right now that this post will run long and there's no need for either of you to read and suffer it if you don't want to. It's written mostly as a project prospectus for daz.
Yep. I was looking at some designs last night and reasoned that the output at the cathode follower (which is the end of the preamp "signal processing" for this amps design) is probably on the order of 60+V. The Soldano design and others attenuate this to some 1/45, or about 1.3V or 1.4V. IME most (not all) signal processors handle this without flinching.
And then there's the issue of impedance.?. The output of the effects loop should be low to avoid signal degradation in the cables. I've tried to ignore this and noted that even 5' of cable through the loop can cause audible signal degradation in the HF with a high impedance output. So we do need a low impedance output to the effects. If the effects loop is to sound transparent (in an amp not designed with one originally) the recovery stage should have a similar output impedance to the place where we cut into the signal path. Short of that the stage following the recovery will see a different impedance relationship and, because that affects dynamics @ frequency, the sound/response/feel of the amp will change. That's been my own limited experience on the matter. Soldano, in the previously mentioned design, even went to the trouble of running the recovery through an additional triode arranged as a cathode follower to achieve this. So... We can get 1.3V from the amp at a low impedance pretty easily. But now we need to recover that and amplify the voltage some 33dB back to 60-ish at a similar output impedance to the cathode follower where we started!!! That's a job for a tube if I ever saw one. I'm no expert, but a corresponding SS design might be unnecessarily difficult by comparison.
I'm actually learning something on this one. I erroneously thought that all the SS effects loops in so many otherwise tube amps were simply using the SS devices as direct substitutes for what a tube would do. Because they're smaller, cheaper and make less heat. So why not use them in any application that isn't tube tone dependent. Made sense to me. But I didn't realize mow much, or the specific design criteria that has to be considered. Now I'm finding myself sort of painted into a corner design wise.
So... At this point I think the best answer is to either add a tube OR tap the effects loop in a location where duplicating the original source voltage will be easier. Padding from cathode follower at some 60+V and recovering to 60+V @ observed relative impedance to feed the tone stack now seem too cumbersome. That is probably why other (seemingly more sane) designs tap at the treble pot output, where the signal will be padded by the existing circuit some 8dB down to a much more manageable 25V and the recovery stage output impedance is less critical. This is just a much more easily achieved goal. In the end this is pretty much just what the Metropoulos module does. So I say now that it is a smarter design than the idea I proposed earlier. It can be bought for $85 bucks or a similar, but as yet undersigned circuit can be built for $10 plus the cost of time ordering parts and a lot of head scratching. I'd buy the Metroamps unit.
Yep. I was looking at some designs last night and reasoned that the output at the cathode follower (which is the end of the preamp "signal processing" for this amps design) is probably on the order of 60+V. The Soldano design and others attenuate this to some 1/45, or about 1.3V or 1.4V. IME most (not all) signal processors handle this without flinching.
And then there's the issue of impedance.?. The output of the effects loop should be low to avoid signal degradation in the cables. I've tried to ignore this and noted that even 5' of cable through the loop can cause audible signal degradation in the HF with a high impedance output. So we do need a low impedance output to the effects. If the effects loop is to sound transparent (in an amp not designed with one originally) the recovery stage should have a similar output impedance to the place where we cut into the signal path. Short of that the stage following the recovery will see a different impedance relationship and, because that affects dynamics @ frequency, the sound/response/feel of the amp will change. That's been my own limited experience on the matter. Soldano, in the previously mentioned design, even went to the trouble of running the recovery through an additional triode arranged as a cathode follower to achieve this. So... We can get 1.3V from the amp at a low impedance pretty easily. But now we need to recover that and amplify the voltage some 33dB back to 60-ish at a similar output impedance to the cathode follower where we started!!! That's a job for a tube if I ever saw one. I'm no expert, but a corresponding SS design might be unnecessarily difficult by comparison.
I'm actually learning something on this one. I erroneously thought that all the SS effects loops in so many otherwise tube amps were simply using the SS devices as direct substitutes for what a tube would do. Because they're smaller, cheaper and make less heat. So why not use them in any application that isn't tube tone dependent. Made sense to me. But I didn't realize mow much, or the specific design criteria that has to be considered. Now I'm finding myself sort of painted into a corner design wise.
So... At this point I think the best answer is to either add a tube OR tap the effects loop in a location where duplicating the original source voltage will be easier. Padding from cathode follower at some 60+V and recovering to 60+V @ observed relative impedance to feed the tone stack now seem too cumbersome. That is probably why other (seemingly more sane) designs tap at the treble pot output, where the signal will be padded by the existing circuit some 8dB down to a much more manageable 25V and the recovery stage output impedance is less critical. This is just a much more easily achieved goal. In the end this is pretty much just what the Metropoulos module does. So I say now that it is a smarter design than the idea I proposed earlier. It can be bought for $85 bucks or a similar, but as yet undersigned circuit can be built for $10 plus the cost of time ordering parts and a lot of head scratching. I'd buy the Metroamps unit.
Comment